| Popis: |
Although widely used in construction and industrial applications, wood is more prone to defects of different kinds than other materials. These defects are unpredictable and differing randomly from plank to plank. This uncertain nature of the defects complicates establishment of manufacturing plans. In this study, a probabilistic model of wood defects was constructed as a function of three variables which were quantity of defects, position of defects and size of defects. The Kolmogorov-Smirnov hypotheses testing on distributional forms of these variables were carried out. Results showed that Poisson, uniform, and log-normal distributions were suitable to represent the variables statistically. Being knowledgeable of how the defects are distributed on the plank will be of benefit in profitability justification of a cutting plan. |
